Genetic diversity and genetic relationship of Jatropha curcas between China and Southeast Asian revealed by amplified fragment length polymorphisms

At present, due to the increasing popularity of Jatropha curcas as a source of biodiesel and the generation of nontoxic and high yielding varieties of the plant, there is an immediate need for genetic improvement. However, conclusions on the genetic basis of J. curcas are still inconsistent. The objective of this study is to use amplified fragment length polymorphisms (AFLP) to survey the genetic diversity of J. curcas in Southeast Asian countries and China to aid in genetic improvement, and to preliminarily get the message on the genetic relationships of J. curcas between China and Southeast Asian. We collected a total of 240 samples from three Asian countries, two African countries and different geographical regions in China. In this study, six AFLP primer combinations were used to amplify, and they yielded a total of 352 scorable loci, 14.78% of which were polymorphic. The number of loci scored per primer combination ranged from 53 (E-ACT/M-CAA and E-ACG/M-CAT) to 68 (E-ACC/MCAA), with an average of 58.67 loci per primer combination. The rate of polymorphisms found for different primer combinations ranged from 12.90 (E-ACA/M-CAT) to 16.98% (E-ACG/M-CAT). The AMOVA revealed that 36.18% of the variance occurred among populations and 63.82% occurred within populations. The above indicators suggest that the genetic germplasms of J. curcas have a narrow genetic diversity in China and Southeast Asia. The clustering of genotypes based on the AFLP markers shows that the genetic relationship between cluster II and III is close, indicating that the origin of J. curcas in China may be from Southeast Asian. Our next work is to find the bands linkage with specific populations or specific properties and convert it into a sequence-characterized amplified region marker to aid in the genetic improvement of this energy plant.

becoming increasingly problematic.Fossil fuels such as oil and natural gas are being used on a large scale, and are the most important energy source.However, these sources will soon be depleted, and deleterious waste products produced by the combustion of fossil fuels may create a number of ecological problems, including the greenhouse effect.Globally, a number of factors, including the increased price of petroleum, the desire to reduce carbon dioxide emissions and fuel security, encourage the search for substitutions for fossil fuels and a reduction of our dependency on crude oil.Bioenergy is a renewable source of primary energy, and its sustainable use does not emit carbon dioxide.The increased use of this energy source may contribute to the achievement of the objectives of the Framework Convention on Climate Change (FCCC), which seek to stabilize atmospheric concentrations of greenhouse gases below dangerous levels (Fischer and Schrattenholzer, 2001;Peters and Thielmann, 2008;King et al., 2009).
Plant oils represent a renewable source of long-chain hydrocarbons that can be used as both fuel and chemical feedstock, and recently, the demand for plant oils as a source of biodiesel in particular has increased.Jatropha curcas has attracted considerable attention as a source of seed-oil (Giibitz et al., 1999;Openshaw, 2000;Jongschaap et al., 2007;Fairless, 2007;Dyer and Mullen, 2008;Kumar and Sharma, 2008).J. curcas belongs to the family of Euphorbiaceae, and was originally grown in countries of the equatorial Americas, but has spread to other tropical countries (Heller, 1996).As a multipurpose plant, Jatropha has many attributes and considerable potential.The plant can be used to prevent and/or control erosion, to reclaim land, grown as a live fence and be planted as a commercial crop.J. curcas seeds are rich in oil and when extracted, pure plant oil can be used directly to produce light, heat and electricity, or it can be used as a feedstock for bio-diesel (Jongschaap et al., 2007;Kumar and Sharma, 2008).
Due to the increasing popularity of J. curcas as a feedstock for biodiesel, generating nontoxic and high yielding varieties of the plant is an immediate need.Many countries have made programs to develop J. curcas for its seed-oil.The success of these programs depends on the identification of genetically divergent material.
At present, reports on the genetic basis of J. curcas came mostly from India and China.For example, Sujatha et al. (2005) determined the similarity index between the toxic Indian and non-toxic Mexican genotypes.Basha and Sujatha (2007) reported modest levels of genetic variation among the 42 accessions of J. curcas from India.Ranade et al. (2008) reported that the wild to semiwild or naturalized accessions of J. curcas from India possess adequate genetic diversity.Tatikonda et al. (2009) reported a broad genetic base of J. curcas in India.Popluechai et al. (2009) assessed genetic variation in 38 J. curcas accessions from 13 countries in three continents, and the results revealed narrow genetic diversity.Overall, these studies demonstrate inconsistent results in examining the genetic diversity of J. curcas.In China, He et al. (2007) reported high levels of genetic diversity among nine populations of J .curcas in China.Xiang et al. (2007) also reported high levels of genetic diversity of J .curcas from only Yunnan province, China.Sun et al. (2008) reported that the genetic variation of J. curcas accessions is very limited in China.Ou et al. (2009) reported high levels of genetic diversity of J. curcas in China.Therefore, we could know that the conclusions on the genetic diversity of J. curcas in China are also inconsistent.
Among different marker systems available, AFLP (Vos et al., 1995) is a highly efficient dominant molecular marker that is stabile and repeatable, and allows the simultaneous analysis of large numbers of marker loci throughout the genome.This marker system has been used to determine genetic diversity in a number of plant species, such as tea (Paul et al., 1997), maize (Marsan et al., 1998) and Cucurbita pepo (Ferriol et al., 2003).J. curcas is widely self-sown and cultivated in southern and southwestern China, but no report has examined the genetic diversity of J. curcas in Southeast Asian countries.We collected materials from three Asian countries, two African countries and different geographical regions in China.We collected seeds and young leaves from the wild, except for those from Mali, Burkina Faso and Indonesia, which were from nurseries.The objective of this study was to use AFLP markers method to survey the genetic diversity of J. curcas in China and Southeast Asia to aid in genetic improvement and preliminarily examine the genetic relationship of J. curcas between China and Southeast Asian countries.

Plant materials
In this study, with the exception of material collected in Mali, Burkina Faso and Indonesia, we were able to collect materials from the wild.A set of 240 samples of J. curcas were collected, including 150 accessions from ten different geographical regions within China (Guizhou, Sichuan, Hainan, Yunnan, Guangxi province).In every province, we selected two different geographic distributions and 90 accessions from five different countries of Asia and Africa.The list of collection sites are shown in Table 1.

Genomic DNA extraction
Genomic DNA was extracted from every accession.J. curcas is rich in polyphenols and polysaccharides, therefore, the quality and quantity of genomic DNA isolated using a traditional CTAB (2%) (Doyle and Doyle, 1990) protocol was poor, and a modified protocol was used as described here.The leaf tissue was ground in liquid nitrogen and placed in a 1.5 ml microcentrifuge tube.To the ground sample, 0.8 ml of extraction buffer (2% CTAB, 100 Mm, Tris-HCl, 1.4 M NaCl, 20 mM EDTA, 0.2 M β-mercaptoethanol, 2% polyvinylpyrrolidone (PVP), pH 8.0) was added and incubated at 65°C for 60 -90 min with occasional gentle shaking.The samples were extracted twice with an equal volume of chloroform: isoamyl alcohol (24:1), and the supernatant was transferred into a new tube.A 2/5 volume of sodium acetate (3 M) was added to the supernatant, and a 2/3 -1 volume of frozen isopropyl alcohol was added to precipitate the DNA.The pelleted DNA was air dried and dissolved in Milli Q water, which contained RNase A (10 mg/ml), and incubated for 30 min at 37°C.
Purification steps were performed as follows: A total of 0.4 ml extraction buffer was added to the tube, and the solution was extracted once with an equal volume of chloroform: isoamyl alcohol (24:1).Finally, the DNA was pelleted with 2/5 volume of sodium acetate (3 M) and one volume of isopropanol, followed by an ethanol wash (70%) twice.The pellet was air dried and resuspended in appropriate volume of low salt tris EDTA (TE).The quantity of the DNA was estimated spectrophotometrically, and the quality was confirmed by 0.8% agarose electrophoresis.

AFLP analysis
AFLP analysis was carried out as described by Vos et al. (1995).The ligated DNA was preamplified with primers (compatible to the EcoRI and MseI adapters) with one selective nucleotide, respectively, in the thermal cycler (Bio-Rad, MC013208), using the following cycling parameters: 30 cycles at 94°C for 30 s, 56°C for 60 s and 72°C for 60 s.The diluted (10-fold) amplified products were used as the template for selective amplification.The second amplification was carried out with six selective primer combinations (Table 2) of EcoRI and MseI with three selective nucleotides.The Touch-down polymerase chain reaction (PCR) program, which is composed of two segments, was performed.PCR was performed using 65°C as the initial annealing temperature for the first cycle and for subsequent 11 cycles the annealing temperature was successively reduced by 0.7°C.Twenty-three cycles were run at 56°C annealing temperature.

Statistical analysis
Using the AFLP polymorphic band statistical method, each allele was scored as present (1) or absent (0) for each loci to create the binary data set.Only the bands of <50 bp in length were scored when clear and repeated.Bands of similar size and intensity were assumed to be homologous.The percentage of polymorphic loci, Nei's gene diversity index (Nei, 1973), Shannon's information index (Lewontin, 1972) and gene flow were calculated using the software, POPGENE, Version 1.31 (Yeh et al., 1997).A Jaccard's similarity coefficient was used to estimate genetic similarity (GS).The GS matrices were subjected to unweighted pair group method analysis (UPGMA) to construct dendrograms with the software, NTSYSPC, Version 2.1 (Rohlf, 2000).The analysis of molecular variance (AMOVA) was also used to partition the total phenotypic variance within and among populations with the software, Alrequin, Version 3.02 (Excoffier and Schneider, 2005).The number of permutations for significant testing was set at 5000 for analysis.

RESULTS
Different primer combinations were screened, and the band numbers and fragment length were different in each primer combination.Six primer combinations were identified.The six AFLP primer pair combinations yielded a total of 352 scorable loci, of which, 14.87% were polymorphic.The number of loci scored per primer combination ranged from 53 (E-ACT/M-CAA and E-ACG/M-CAT) to 68 (E-ACC/M-CAA), with an average of 58.67 loci per primer combination.The rate of polymorphism found for different primer combinations ranged from 12.90 (E-ACA/M-CAT) to 16.98% (E-ACG/M-CAT) (Table 2).Nei's gene diversity ranged from 0.0460 (E-ACT/M-CAA) to 0.0669 (E-ACG/M-CAT), the Shannon's information index ranged from 0.071 (E-ACT/M-CAA) to 0.0984 (E-ACG/M-CAT) and the effective number of alleles ranged from 1.0733 (E-ACT/M-CAA) to 1.1116 (E-ACG/M-CAT) (Table 3).To assess the overall distribution of diversity within and among these populations, an AMOVA analysis was completed (Table 4).AMOVA revealed that 36.18% of the variance occurred among populations and 63.82% within populations.The above indicators suggest that the genetic germplasm of J. curcas possesses a narrow genetic diversity.A dendrogram (Figure 1) constructed using data from the UPGMA cluster analysis, based on the genetic similarity coefficient matrix of all the populations analyzed, illustrates the overall genetic relationship among the genotypes surveyed.The sixteen populations from different geographical regions were divided into three major clusters at a coefficient of genetic similarity: Cluster I included two populations (Mai and Burkina Faso, Africa), Cluster II comprised of four populations from Southeast Asia (Laos, Burma and two Indonesia populations) and Cluster III comprised of the other populations from China.

DISCUSSION
AFLP is an information rich marker system with the ability to generate a large number of polymorphic/informative loci simultaneously in a single lane with a single-primer combination, compared with random amplified polymerphic DNA (RAPD), restriction fragment length polymorphism (RFLP) and microsatellites (Powell, 1996;Milbourne et al., 1997;Russell et al., 1997).AFLP can reveal the complete hereditary feature of germplasm resources, especially those with a close genetic relationship, and is a powerful approach for the assessment of genetic variation among populations.
From the introduction, we have known these reported results formally, in which the genetic diversity of J. curcas at home and abroad is non-uniform.This may be due to the use of different grow-status (wild, semi-wild or cultivated) materials or different molecular marker methods.In this study, we adopted the AFLP method to reveal the genetic diversity and our results showed that the genetic diversity of the J. curcas examined is narrow, and is consistent with previous reports (Sun et al., 2008;Popluechai et al., 2009).However, it is worth mentioning that the sources of J. curcas used by Sun et al. (2008) and the accessions were collected and propagated by seeds in an open field in a botanical garden.J. curcas is monoecious, and normally, the flowers are unisexual, such that male and female flowers are produced in the same inflorescence (Li et al., 2007;Liu et al., 2008).In addition, Li et al. (2007), Yang et al. (2007) and Luo et al. (2007) researched the pollination biology of J. curcas, and showed a tendency to promote xenogamy and  2008) is questionable to some extent.In this study, we tried our best to collect materials from the wild, except for Mali, Burkina Faso and Indonesia, which were sampled from nurseries.
The 16 populations that came from different geographical regions were divided into three major clusters (Figure 1).In Figure 1, the two populations from Africa are distant from the other populations, demonstrating that the genetic relationship of the populations between Asia and African is distant, which may be due to geographical separation.The genetic relationship between cluster II and III is close, indicating that the source of J. curcas in China may be from Southeast Asian countries.In the history of Asia, India was the earliest Asian country that Portuguese seafarers arrived.Therefore, we speculate that the original source of J. curcas in Asia was in India, and may be introduced by Portuguese colonists or other people.Geographically, the Indian subcontinent is adjacent to the Southeast Asia, so J. curcas were spread to Southeast Asian countries (e.g.Burma) from land or were diffused to islands (e.g.Indonesia) in Southeast Asian countries by Portuguese colonists.Yunnan and Guangxi provinces, China, are adjacent to Burma and Vietnam, respectively.However, because of the sources of materials and the adopted research method, the conjecture that the original source of J. curcas in Asia is in India, is not supported by our results.In future experiments, more material from different areas, including from India, Southeast Asian countries and China, should be collected, and other molecular marker methods should be used.The reasons for low genetic diversity of J. curcas in China may be: J. curcas belongs to the family of Euphorbiaceae and originally grew in countries of the equatorial Americas, from where it has been spread by Portuguese seafarers to other tropical countries (Heller, 1996).When J. curcas spread to a region, then the center around the ground to spread is the perimeter zone.The original source of J. curcas in Asia may be in India, and may be introduced by the Portuguese colonists, and the time is probably about five hundred years.Although J. curcas is widely distributed around the world now, its source was limited as a result of genetic causes in some specific habitat environment with different altitude, latitude, etc., which may be due to the fact that the numbers of plant introduced in India were small.In addition, the time in which it was introduced into Asia was not long enough to give rise to the limited genetic variation, so J. curcas in China has a low genetic diversity, just as the whole of Asia has a low genetic diversity.But this conclusion is limited to China and Southeast Asia, other parts of J. curcas can not determine the genetic basis.The low genetic basis and vegetative propagation will result in limited genetic variation among species in regions where it has been introduced.Basha and Sujatha (2007) attributed the limited genetic base of J. curcas in India to the small number of introduced plants and their vegetative propagation.Sun et al. (2008) suggested that the genetic diversity of J. curcas in China may represent only part of the Indian population.This may explain the cause of the narrow genetic diversity of J. curcas in China.
The AMOVA revealed that 36.18% of the variance occurred among populations and 63.82% occurred within population.J. curcas has a wider distribution range, resulting to the different geographical environments in different populations.These geographic environments have different ecological factors such as temperature, moisture, light and so on, that result in different populations being exposed to quite different habitat environment selection pressure, eventually leading to genetic differentiation among populations.In addition, due to the limited transmission distance of pollen and since birds do not eat the toxic seeds, the characteristics are not conducive to long-distance spread, which may lead to a blocked gene flow among populations (Xiang et al., 2007;Ou et al., 2009).
At present, due to the increasing popularity of J. curcas as a feedstock for biodiesel, there is an immediate need to generate nontoxic and high yielding varieties of the plant through genetic improvement.However, the low level of genetic variation is likely to limit progress in any genetic improvement program for this energy crop.Therefore, we need to increase the limited genetic base of J. curcas.Although the genetic diversity of J. curcas in its native Central America is unknown, germplasm for genetic improvement programs should be introduced from its native range rather than from other introduced populations (Sun et al., 2008).This suggests that there is a realistic and imperative need to determine the genetic basis of J. curcas in its primitive habitat.Despite the limited genetic diversity within J. curcas, appreciable variability exists in important phenotypic, physiological and biochemical traits such as seed size, water use efficiency and seed oil content, respectively (Akintayo, 2004;Kaushik et al., 2007;Berchmans and Hirata, 2008).This variability suggests differences in fundamental epigenetic regulatory mechanisms and implies J. curcas as a unique system to study genetic variability (Popluechai et al., 2009).
On the other hand, the genus Jatropha is morphologically diverse, encompassing more than 200 species that are distributed chiefly in the dry tropical regions of America, and were introduced into Africa and Asia and are now cultivated worldwide (Sujatha and Prabakaran, 1997;Fairless, 2007).Some reports of the genetic diversity between some species in Jatropha genus have demonstrated a high level of genetic variation among the Jatropha species (Ram et al., 2008;Pamidimarri et al., 2008;Kumar et al., 2008;Basha and Sujatha, 2009;Popluechai et al., 2009).High-level genetic variation among species of Jatropha suggests that we may be able to work on genetic improvement to J. curcas by interspecific hybrids.A naturally occurring interspecific hybrid, Jatropha tanjorensis Ellis × Saroja, had been identified in India by Prabakaran and Sujatha (1999).Interspecific breeding (J.curcas × J. integerrima) has been achieved successfully (Sujatha and Prabakaran, 2003).Popluechai et al. (2009) have obtained initial success in generating low phorbol esters by interspecific hybrids.In addition, the genetic improvement of this energy plant also can be possible by genetic transfor-mation.Li et al. (2008) have established an Agrobacterium-mediated cotyledon disc transformation method for J. curcas, and He et al. (2009) have established a method of Agrobacterium tumefaciensmediated transformation of the calli of J. curcas.
At present, some genes of J. curcas have been cloned and expressed (Lin et al., 2003;Li et al., 2008).In addition, with the construction of gene libraries and genome sequencing that is underway (ISAAA, 2008), the course genetic improvement of J. curcas may be expedited greatly.J. curcas may be a useful and unique system to further define the molecular mechanisms on regulatory or prominent phenotypic variability.The markers generated by AFLP can be employed efficiently in marker assisted selection (MAS) and genetic resource management.Our next work is to identify band linkage with specific populations or specific properties, and convert these results into SCAR markers to help for genetic improvement of this energy plant, collect larger amounts of material and adopt new methods to affirm the provenance of J. curcas in China.

Figure 1 .
Figure 1.The dendrogram generated by AFLP data using the UPGMA method of J. curcas.

Table 1 .
Source locations for the different accessions of J. curcas.

Table 2 .
The six primer combinations used for selective amplification and the results amplified by it.
The appropriate genomic DNA was digested with restriction enzymes, EcoRI and MseI (New England Biolabs, USA), at 37°C for 4 h.EcoRI and MseI specific adapters were ligated to the restricted DNA fragments at 16°C overnight.

Table 3 .
Estimates of genetic diversity among J. curcas.

Table 4 .
Analysis of molecular variance in 16 populations of J. curcas.